Help:Extension:ParserFunctions/en

<languages/> The Template:Ll extension provides additional parser functions to supplement the "Template:Ll", which are already present in MediaWiki. (It may be configured to provide additional parser functions for string handling; these string functions are documented Template:Ll.) All the parser functions provided by this extension take the form:

{{#functionname: argument 1 | argument 2 | argument 3 ... }}

#expr

Template:Hl1 | Type	Template:Hl1 | Operators
Template:Hl2 | Grouping (parentheses)	( )
Template:Hl2 | Numbers	1234.5   e (2.718)   pi (3.142)
Template:Hl2 |	binary operator e   unary +,-
Template:Hl2 | Unary	not ceil trunc floor abs exp ln sin cos tan acos asin atan sqrt
Template:Hl2 rowspan="3" | Binary	^
* / div mod fmod
+ -
Template:Hl2 | Round	round
Template:Hl2 rowspan="3" | Logic	= != <> > < >= <=
and
or

This function evaluates a mathematical expression and returns the calculated value. This function is also available in Template:Ll via the mw.ext.ParserFunctions.expr function.

{{#expr: expression }}

Basic example

{{#expr: 1 + 1 }} → 2

The available operators are listed to the right, in order of precedence. See Template:Ll for more details of the function of each operator. The accuracy and format of the result returned will vary depending on the operating system of the server running the wiki and the number format of the site language.

When evaluating using boolean algebra, zero evaluates to Template:Phpi, and any nonzero value, positive or negative, evaluates to Template:Phpi:

{{#expr: 1 and -1 }} → 1

{{#expr: 1 and 0 }} → 0

{{#expr: 1 or -1 }} → 1

{{#expr: -1 or 0 }} → 1

{{#expr: 0 or 0 }} → 0

An empty input expression returns an empty string. Invalid expressions return one of several error messages, which can be caught using the #iferror function:

{{#expr: }} →

{{#expr: 1+ }} →

{{#expr: 1 = }} →

{{#expr: 1 foo 2 }} →

The order of addition and subtraction operands before or after a number is meaningful and may be treated as a positive or negative value instead of as an operand with an erroneous input:

{{#expr: +1 }} → 1

{{#expr: -1 }} → -1

{{#expr: + 1 }} → 1

{{#expr: - 1 }} → -1

Note, if using the output of magic words, you must raw-format them in order to remove commas and translate the numerals. For example, {{NUMBEROFUSERS}} results in 0, where we want 0, which can be obtained using {{Template:Ll:{{NUMBEROFUSERS}}|R}}. This is especially important in some languages, where numerals are translated. For example, in Bengali, {{NUMBEROFUSERS}} produces ৩০,০৬১.

{{#expr:{{NUMBEROFUSERS}}+100}} → 100

{{#expr:{{formatnum:{{NUMBEROFUSERS}}|R}}+100}} → 100

Rounding

Rounds off the number on the left to a multiple of 1/10 raised to a power, with the exponent equal to the truncated value of the number given on the right.

To round up or down use unary ceil or floor respectively.

Template:Hl1 | Test case	Template:Hl1 | Result	Template:Hl1 | Method of rounding
{{#expr: 1/3 round 5 }}	0.33333	Final digit is < 5, so no apparent rounding occurs (0.333333… → 0.33333)
{{#expr: 1/6 round 5 }}	0.16667	Final digit is ≥ 5, so it is rounded up (0.166666… → 0.16667)
{{#expr: 8.99999/9 round 5 }}	1	Again, the result is rounded up on the last digit, which results in additional rounding (0.999998… → 1.00000 → 1)
{{#expr: 1234.5678 round -2 }}	1200	Rounded to nearest 100 because negative values round to the left of the decimal point
{{#expr: 1234.5678 round 2 }}	1234.57	Rounded to nearest 100th because positive values round to the right of the decimal point
{{#expr: 1234.5678 round 2.3 }}	1234.57	Decimals in the rounding index make no difference in the rounded result
{{#expr: trunc 1234.5678 }}	1234	Decimal portion truncated (chopped off)
colspan=3 Template:Hl2 | Rounding to the nearest integer
{{#expr: 1/3 round 0 }}	0	Down to the nearest integer, which is zero
{{#expr: 1/2 round 0 }}	1	Up to the nearest integer, which is one
{{#expr: 3/4 round 0 }}	1	Up to the nearest integer, which is one
{{#expr: -1/3 round 0 }}	-0	Up to the nearest integer, which is zero
{{#expr: -1/2 round 0 }}	-1	Down to the nearest integer, which is negative one
{{#expr: -3/4 round 0 }}	-1	Down to the nearest integer, which is negative one
colspan=3 Template:Hl2 | Rounding up or down with ceil and floor
{{#expr: ceil(1/3) }}	1	Up to the next larger integer, which is one
{{#expr: floor(1/3) }}	0	Down to the next smaller integer, which is zero
{{#expr: ceil(-1/3) }}	-0	Up to the next larger integer, which is zero
{{#expr: floor(-1/3) }}	-1	Down to the next smaller integer, which is negative one
{{#expr: ceil 1/3 }}	0.33333333333333	Not rounded, since 1 already is an integer Template:Warning
colspan=3 Template:Hl2 | Rounding large numbers
{{#expr: 1e-92 round 400 }}	1.0E-92	Rounding to a very large number leads to infinity. Hence, the original value without the infinity is given as the answer.
{{#expr: 1e108 round 200 }}	1.0E+108	Same as above.

Strings

Expressions only work with number-like values, they cannot compare strings or characters. #ifeq can be used instead.

{{#expr: "a" = "a" }} →

{{#expr: a = a }} →

{{#ifeq: a | a | 1 | 0 }} → 1

#if

This function evaluates a test string and determines whether or not it is empty. A test string containing only white space is considered to be empty.

{{#if: test string | value if test string is not empty | value if test string is empty (or only white space)}}

{{#if: first parameter | second parameter | third parameter }}

This function first tests whether the first parameter is not empty. If the first parameter is not empty, the function displays the second argument. If the first parameter is empty or contains only whitespace characters (spaces, newlines, etc.) it displays the third argument.

{{#if: | yes | no}} → no

{{#if: string | yes | no}} → yes

{{#if:      | yes | no}} → no

{{#if: | yes | no}} → no

The test string is always interpreted as pure text, so mathematical expressions are not evaluated (see #ifexpr for that):

{{#if: 1==2 | yes | no }} → yes

{{#if: 0 | yes | no }} → yes

The last parameter (false) may be omitted:

{{#if: foo | yes }} → yes

{{#if: | yes }} →

{{#if: foo | | no}} →

The function may be nested. To do so, nest the inner #if function in its full form in place of a parameter of the enclosing #if function. Up to seven levels of nesting is possible, although that may depend on the wiki or a memory limit.

{{#if:test string
|value if test string is not empty
|{{#if:test string
  |value if test string is not empty
  |value if test string is empty (or only white space)
  }}
}}

You can also use a parameter as the test string in your #if statement. You need to ensure you add the | (pipe symbol) after the name of the variable. (So that if the parameter does not have a value, it evaluates to an empty string instead of the string "{{{1}}}".)

Template:Tmpl

See Template:Ll for more examples of this parser function.

#ifeq

This parser function compares two input strings, determines whether they are identical, and returns one of two strings based on the result. If more comparisons and output strings are required, consider using #switch.

{{#ifeq: string 1 | string 2 | value if identical | value if different }}

If both strings are valid numerical values, the strings are compared numerically:

{{#ifeq: 01 | 1 | equal | not equal}} → equal

{{#ifeq: 0 | -0 | equal | not equal}} → equal

{{#ifeq: 1e3 | 1000 | equal | not equal}} → equal

{{#ifeq: {{#expr:10^3}} | 1000 | equal | not equal}} → equal

Otherwise, the comparison is made as text; this comparison is case-sensitive:

{{#ifeq: foo | bar | equal | not equal}} → not equal

{{#ifeq: foo | Foo | equal | not equal}} → not equal

{{#ifeq: "01" | "1" | equal | not equal}} → not equal  (compare to similar example above, without the quotes)

{{#ifeq: 10^3 | 1000 | equal | not equal}} → not equal  (compare to similar example above, with #expr returning a valid number first)

As a practical example, consider an existing Template:Ll Template:Timer using the parser to choose between two standard times, short and long. It takes the parameter as the first input to compare against the string "short" – there is no convention for the order, but it is simpler to read if the parameter goes first. The template code is defined as:

{{#ifeq: {{{1|}}} | short | 20 | 40 }}

the following ensue:

{{timer|short}} → 20

{{timer|20}} → 40

{{timer}} → 40

#iferror

This function takes an input string and returns one of two results; the function evaluates to true if the input string contains an HTML object with class="error", as generated by other parser functions such as #expr, #time and #rel2abs, template errors such as loops and recursions, and other "failsoft" parser errors.

{{#iferror: test string | value if error | value if correct }}

One or both of the return strings can be omitted. If the correct string is omitted, the test string is returned if it is not erroneous. If the error string is also omitted, an empty string is returned on an error:

{{#iferror: {{#expr: 1 + 2 }} | error | correct }} → correct

{{#iferror: {{#expr: 1 + X }} | error | correct }} → error

{{#iferror: {{#expr: 1 + 2 }} | error }} → 3

{{#iferror: {{#expr: 1 + X }} | error }} → error

{{#iferror: {{#expr: 1 + 2 }} }} → 3

{{#iferror: {{#expr: 1 + X }} }} → ‌

{{#iferror: {{#expr: . }} | error | correct }} → correct

{{#iferror: <strong class="error">a</strong> | error | correct }} → error

Some errors may cause a tracking category to be added, using {{#iferror:}} will not suppress the addition of the category.

#ifexpr

This function evaluates a mathematical expression and returns one of two strings depending on the boolean value of the result:

{{#ifexpr: expression | value if true | value if false }}

The expression input is evaluated exactly as for #expr above, with the same operators being available. The output is then evaluated as a boolean expression.

An empty input expression evaluates to Template:Phpi:

{{#ifexpr: | yes | no}} → no

As mentioned above, zero evaluates to Template:Phpi and any nonzero value evaluates to Template:Phpi, so this function is equivalent to one using #ifeq and #expr only:

{{#ifeq: {{#expr: expression }} | 0 | value if false | value if true }}

except for an empty or wrong input expression (an error message is treated as an empty string; it is not equal to zero, so we get value if true).

{{#ifexpr: = | yes | no }} →

comparing

{{#ifeq: {{#expr: = }} | 0 | no | yes }} → yes

Either or both of the return values may be omitted; no output is given when the appropriate branch is left empty:

{{#ifexpr: 1 > 0 | yes }} → yes

{{#ifexpr: 1 < 0 | yes }} →

{{#ifexpr: 0 = 0 | yes }} → yes

{{#ifexpr: 1 > 0 | | no}} →

{{#ifexpr: 1 < 0 | | no}} → no

{{#ifexpr: 1 > 0 }} →

Boolean operators of equality or inequality operators are supported.

{{#ifexpr: 0 = 0 or 1 = 0 | yes}} → yes

{{#ifexpr: 0 = 0 and 1 = 0 | | no}} → no

{{#ifexpr: 2 > 0 or 1 < 0 | yes}} → yes

{{#ifexpr: 2 > 0 and 1 > 0 | yes | no}} → yes

#ifexist

This function takes an input string, interprets it as a page title, and returns one of two values depending on whether or not the page exists on the local wiki.

{{#ifexist: page title | value if exists | value if doesn't exist }}

The function evaluates to Template:Phpi if the page exists, whether it contains content, is visibly blank (contains meta-data such as category links or Template:Ll, but no visible content), is blank, or is a Template:Ll. Only pages that are redlinked evaluate to Template:Phpi, including if the page used to exist but has been deleted.

{{#ifexist: Help:Extension:ParserFunctions/en | exists | doesn't exist }} → exists

{{#ifexist: XXHelp:Extension:ParserFunctions/enXX | exists | doesn't exist }} → doesn't exist

The function evaluates to Template:Phpi for Template:Ll that have been customized, and for Template:Ll that are defined by the software.

{{#ifexist: Special:Watchlist | exists | doesn't exist }} → exists

{{#ifexist: Special:CheckUser | exists | doesn't exist }} → doesn't exist (because the Template:Ll extension is not installed on this wiki)

{{#ifexist: MediaWiki:Copyright | exists | doesn't exist }} → doesn't exist (because MediaWiki:Copyright has not been customized)

If a page checks a target using #ifexist:, then that page will appear in the Special:WhatLinksHere list for the target page. So if the code {{#ifexist:Foo }} were included live on this page (Help:Extension:ParserFunctions/en), Special:WhatLinksHere/Foo will list Help:Extension:ParserFunctions/en.

On wikis using a shared media repository, #ifexist: can be used to check if a file has been uploaded to the repository but not to the wiki itself:

{{#ifexist: File:Example.png | exists | doesn't exist }} → doesn't exist

{{#ifexist: Image:Example.png | exists | doesn't exist }} → doesn't exist

{{#ifexist: Media:Example.png | exists | doesn't exist }} → exists

If a local description page has been created for the file, the result is exists for all of the above.

#ifexist: does not work with interwiki links.

ifexist limits

#ifexist: is considered an "expensive parser function"; only a limited number of which can be included on any one page (including functions inside transcluded templates). When this limit is exceeded, any further #ifexist: functions automatically return false, whether the target page exists or not, and the page is categorized into Category:Pages with too many expensive parser function calls. The name of the Template:Ll may vary depending on the content language of your wiki.

For some use cases it is possible to emulate the ifexist effect with css, by using the selectors a.new (to select links to unexisting pages) or a:not(.new) (to select links to existing pages). Furthermore, since the number of expensive parser functions that can be used on a single page is controlled by Template:Ll, one can also increase the limit in LocalSettings.php if needed.

ifexist and wanted pages

A page that does not exist and is tested for using #ifexist will end up on the Wanted Pages. See Template:Phab for the reason, and w:Template:Linkless exists for a workaround.

#rel2abs

This function converts a relative file path into an absolute filepath.

{{#rel2abs: path }}

{{#rel2abs: path | base path }}

Within the path input, the following syntax is valid:

• . → the current level
• .. → go up one level
• /foo → go down one level into the subdirectory /foo

If the base path is not specified, the full page name of the page will be used instead:

{{#rel2abs: /quok | Help:Foo/bar/baz }} → Help:Foo/bar/baz/quok

{{#rel2abs: ./quok | Help:Foo/bar/baz }} → Help:Foo/bar/baz/quok

{{#rel2abs: ../quok | Help:Foo/bar/baz }} → Help:Foo/bar/quok

{{#rel2abs: ../. | Help:Foo/bar/baz }} → Help:Foo/bar

Invalid syntax, such as /. or /./, is ignored. Since no more than two consecutive full stops are permitted, sequences such as these can be used to separate successive statements:

{{#rel2abs: ../quok/. | Help:Foo/bar/baz }} → Help:Foo/bar/quok

{{#rel2abs: ../../quok | Help:Foo/bar/baz }} → Help:Foo/quok

{{#rel2abs: ../../../quok | Help:Foo/bar/baz }} → quok

{{#rel2abs: ../../../../quok | Help:Foo/bar/baz }} →

For a similar group of functions see also Help:Magic words#URL data. Built-in parser functions include: 'localurl:', 'fullurl:', 'anchorencode:' etc.

#switch

See also: w:Help:Switch parser function

This function compares one input value against several test cases, returning an associated string if a match is found.

Examples:

{{#switch: baz | foo = Foo | baz = Baz | Bar }} → Baz

{{#switch: foo | foo = Foo | baz = Baz | Bar }} → Foo

{{#switch: zzz | foo = Foo | baz = Baz | Bar }} → Bar

#switch with partial transclusion tags can affect a configuration file that enables an editor unfamiliar with template coding to view and edit configurable elements.

Default

The default result is returned if no case string matches the comparison string:

{{#switch: test | foo = Foo | baz = Baz | Bar }} → Bar

In this syntax, the default result must be the last parameter and must not contain a raw equals sign (an equals sign without {{}}). If it does, it will be treated as a case comparison, and no text will display if no cases match. This is because the default value has not been defined (is empty). If a case matches however, its associated string will be returned.

{{#switch: test | Bar | foo = Foo | baz = Baz }} →

{{#switch: test | foo = Foo | baz = Baz | B=ar }} →

{{#switch: test | test = Foo | baz = Baz | B=ar }} → Foo

Alternatively, the default result may be explicitly declared with a case string of "#default".

Default results declared in this way may be placed anywhere within the function:

{{#switch: test | foo = Foo | #default = Bar | baz = Baz }} → Bar

If the default parameter is omitted and no match is made, no result is returned:

{{#switch: test | foo = Foo | baz = Baz }} →

Grouping results

It is possible to have 'fall through' values, where several case strings return the same result string. This minimizes duplication.

Here cases 2, 3 and 4 all return result234; cases 6 and 7 both return result67. The "#default = " in the last parameter may be omitted in the above case.

Use with parameters

The function may be used with parameters as the test string. In this case, it is not necessary to place the pipe after the parameter name, because it is very unlikely that you will choose to set a case to be the string "{{{parameter name}}}". (This is the value the parameter will default to if the pipe is absent and the parameter doesn't exist or have a value. See Template:Ll.)

{{#switch: {{{1}}} | foo = Foo | baz = Baz | Bar }}

In the above case, if {{{1}}} equals foo, the function will return Foo. If it equals baz, the function will return Baz. If the parameter is empty or does not exist, the function will return Bar.

As in the section above, cases can be combined to give a single result.

{{#switch: {{{1}}} | foo | zoo | roo = Foo | baz = Baz | Bar }}

Here, if {{{1}}} equals foo, zoo or roo, the function will return Foo. If it equals baz, the function will return Baz. If the parameter is empty or does not exist, the function will return Bar.

Additionally, the default result can be omitted if you do not wish to return anything if the test parameter value does not match any of the cases.

{{#switch: {{{1}}} | foo = Foo | bar = Bar }}

In this case, the function returns an empty string unless {{{1}}} exists and equals foo or bar, in which case it returns Foo or Bar, respectively.

This has the same effect as declaring the default result as empty.

{{#switch: {{{1}}} | foo | zoo | roo = Foo | baz = Baz | }}

If for some reason you decide to set a case as "{{{parameter name}}}", the function will return that case's result when the parameter doesn't exist or doesn't have a value. The parameter would have to exist and have a value other than the string "{{{parameter name}}}" to return the function's default result.

(when {{{1}}} doesn't exist or is empty):

{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }} → Foo

(when {{{1}}} has the value "test"):

{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }} → Bar

(when {{{1}}} has the value "{{{1}}}"):

{{#switch: {{{1}}} | {{{1}}} = Foo | baz = Baz | Bar }} → Foo

In this hypothetical case, you would need to add the pipe to the parameter ({{{1|}}}).

Comparison behavior

As with #ifeq, the comparison is made numerically if both the comparison string and the case string being tested are numeric; or as a case-sensitive string otherwise:

{{#switch: 0 + 1 | 1 = one | 2 = two | three}} → three

{{#switch: {{#expr: 0 + 1}} | 1 = one | 2 = two | three}} → one

{{#switch: 02 | +1 = one | +2 = two | three}} → two

{{#switch: 100 | 1e1 = ten | 1e2 = hundred | other}} → hundred

{{#switch: a | a = A | b = B | C}} → A

{{#switch: A | a = A | b = B | C}} → C

A case string may be empty:

{{#switch: | = Nothing | foo = Foo | Something }} → Nothing

Once a match is found, subsequent cases are ignored:

{{#switch: b | f = Foo | b = Bar | b = Baz | }} → Bar

Raw equal signs

"Case" strings cannot contain raw equals signs. To work around this, use the [[Special:MyLanguage/Help:Magic words#Other|Template:((=Template:))]] magic word, or replace equals sign with HTML code =.

Example:

You type	You get
{{#switch: 1=2 | 1=2 = raw | 1<nowiki>=</nowiki>2 = nowiki | 1{{=}}2 = template | default }}	template
{{#switch: 1=2 | 1&#61;2 = html | default }}	html

Replacing #ifeq

#switch can be used to reduce expansion depth.

For example:

• {{#switch:{{{1}}} |condition1=branch1 |condition2=branch2 |condition3=branch3 |branch4}}

is equivalent to

• {{#ifeq:{{{1}}}|condition1 |branch1 |{{#ifeq:{{{1}}}|condition2 |branch2 |{{#ifeq:{{{1}}}|condition3 |branch3 |branch4}}}}}}

i.e. deep nesting, linear:

{{#ifeq:{{{1}}}|condition1
  |<!--then-->branch1
  |<!--else-->{{#ifeq:{{{1}}}|condition2
                |<!--then-->branch2
                |<!--else-->{{#ifeq:{{{1}}}|condition3
                              |<!--then-->branch3
                              |<!--else-->branch4}}}}}}

On the other hand, the switch replacement could be complicated/impractical for IFs nested in both branches (shown with alternatives of indentation, indented on both sides), making full symmetrical tree:

{{#ifeq:{{{1}}}|condition1
 |<!--then-->branch1t{{
  #ifeq:{{{1}}}|condition2
   |<!--then-->branch1t2t{{#ifeq:{{{1}}}|condition4|<!--then-->branch1t2t4t|<!--else-->branch1t2t4e}}
   |<!--else-->branch1t2e{{#ifeq:{{{1}}}|condition5|<!--then-->branch1t2e5t|<!--else-->branch1t2e5e}}
  }}
 |<!--else-->branch1e{{#ifeq:{{{1}}}|condition3
   |<!--then-->branch1e3t{{#ifeq:{{{1}}}|condition6|branch1e3t6t|branch1e3t6e}}
   |<!--else-->branch1e3e{{
    #ifeq:{{{1}}}|condition7
     |branch1e3e7t
     |branch1e3e7t
    }}
  }}
}}

#time

Template:Hl1 | Code	Template:Hl1 | Description	Template:Hl1 | Current output
Template:Hl2 colspan="3"| Year
Y	4-digit year.	2026
y	2-digit year.	26
L	1 if it's a leap year, 0 if not.	0
o	ISO-8601 year of the specified week.	2026
Template:Hl2 colspan="3"| Month
n	Month index, not zero-padded.	2
m	Month index, zero-padded.	02
M	An abbreviation of the month name, in the site language.	Feb
F	The full month name in the site language.	February
xg	Output the full month name in the genitive form for site languages that distinguish between genitive and nominative forms. This option is useful for many Slavic languages like Polish, Russian, Belarusian, Czech, Slovak, Slovene, Ukrainian, etc.	For Polish: {{#time:F Y|June 2010|pl}} → czerwiec 2010 (nominative) {{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010 (genitive)
Template:Hl2 colspan="3"| Day of the month or the year
j	Day of the month, not zero-padded.	6
d	Day of the month, zero-padded.	06
z	Day of the year (January 1 = 0). Template:Note To get the ISO day of the year add 1.	36
Template:Hl2 colspan="3"| Week and day of the week
W	ISO 8601 week number, zero-padded.	06
N	ISO 8601 day of the week (Monday = 1, Sunday = 7).	5
w	Number of the day of the week (Sunday = 0, Saturday = 6).	5
D	An abbreviation for the day of the week. Rarely internationalized.	Fri
l	The full weekday name. Rarely internationalized.	Friday
Template:Hl2 colspan="3"| Hour
a	"am" during the morning (00:00:00 → 11:59:59), "pm" otherwise (12:00:00 → 23:59:59).	pm
A	Uppercase version of a above.	PM
g	Hour in 12-hour format, not zero-padded.	1
h	Hour in 12-hour format, zero-padded.	01
G	Hour in 24-hour format, not zero-padded.	13
H	Hour in 24-hour format, zero-padded.	13
Template:Hl2 colspan="3"| Minutes and seconds
i	Minutes past the hour, zero-padded.	35
s	Seconds past the minute, zero-padded.	33
U	Unix time. Seconds since January 1 1970 00:00:00 GMT.	1770384933
id="TimeZone" Template:Hl2 colspan="3"| Timezone (as of 1.22wmf2)
e	Timezone identifier.	UTC
I	Whether or not the date is in daylight savings time.	0
O	Difference to Greenwich time (GMT)	+0000
P	Difference to Greenwich time (GMT), with colon	+00:00
T	Timezone abbreviation.	UTC
Z	Timezone offset in seconds.	0
Template:Hl2 colspan="3"| Miscellaneous
t	Number of days in the current month.	28
c	ISO 8601 formatted date, equivalent to Y-m-d"T"H:i:s+00:00.	2026-02-06T13:35:33+00:00
r	RFC 5322 formatted date, equivalent to D, j M Y H:i:s +0000, with weekday name and month name not internationalized.	Fri, 06 Feb 2026 13:35:33 +0000
Template:Hl2 colspan="3" | Non-Gregorian calendars
Template:Hl3 colspan="3" | Islamic
xmj	Day of the month.	18
xmF	Full month name.	Sha'aban
xmn	Month index.	8
xmY	Full year.	1447
Template:Hl3 colspan="3" | Iranian (Jalaly)
xit	Number of days in the month.	30
xiz	Day of the year.	322
xij	Day of the month.	17
xiF	Full month name.	Bahman
xin	Month index.	11
xiY	Full year.	1404
xiy	2-digit year.	04
Template:Hl3 colspan="3" | Hebrew
xjj	Day of the month.	19
xjF	Full month name.	Shevat
xjt	Number of days in month.	30
xjx	Genitive form of the month name.	Shevat
xjn	Month number.	5
xjY	Full year.	5786
Template:Hl3 colspan="3" | Thai solar
xkY	Full year in Thai solar calendar. Template:Note For years before 1941 the dates in Jan-Mar range are not calculated properly.	2569
Template:Hl3 colspan="3" | Minguo/Juche year
xoY	Full year.	115
Template:Hl3 colspan="3" | Japanese nengo
xtY	Full year.	令和8
Template:Hl2 colspan="3" | Flags
xn	Format the next numeric code as a raw ASCII number.	In the Hindi language, {{#time:H, xnH}} produces ०६, 06.
xN	Like xn, but as a toggled flag, which endures until the end of the string or until the next appearance of xN in the string.
xr	Format the next number as a roman numeral. Only works for numbers up to 10,000 (up to 3,000 in pre MediaWiki 1.20).	{{#time:xrY}} → MMXXVI
xh	Format the next number as a Hebrew numeral.	{{#time:xhY}} → ב'כ"ו

This parser function takes a date and/or time (in the Gregorian calendar) and formats it according to the syntax given. A date/time object can be specified; the default is the value of the magic word {{CURRENTTIMESTAMP}} – that is, the time the page was last rendered into HTML.

{{#time: format string }}

{{#time: format string | date/time object }}

{{#time: format string | date/time object | language code }}

{{#time: format string | date/time object | language code | local }}

The list of accepted formatting codes is given in the table to the right. Any character in the formatting string that is not recognized is passed through unaltered; this applies also to blank spaces (the system does not need them for interpreting the codes). If no character is recognized in the formatting string, and the date/time object is without error, then the formatting string is returned as output. There are also two ways to escape characters within the formatting string:

1. A backslash followed by a formatting character is interpreted as a single literal character
2. Characters enclosed in double quotes are considered literal characters, and the quotes are removed.

In addition, the digraph xx is interpreted as a single literal "x".

As the list of formatting codes continues to evolve (with the support of new calendars, or of new date fields computed and formatted differently), you should escape all literal characters (not just ASCII letters currently used by formatting codes) that need to be passed through unaltered.

Unfortunately, for now, the ASCII single quote is still not recognized as a simple alternative for marking literal text to the currently supported ASCII double quotes (for example, double quotes are mandatory for in other uses like the delimitation of string values in JSON, C, C++...) and backslashes (which have to be escaped as well in string constants used by many languages, including JSON, C, C++, PHP, JavaScript, Lua). So you still cannot embed any literal double quote without escaping it with a backslash (or you can use other curly, angular or square quotation marks instead).

{{#time: Y-m-d }} → 2026-02-06

{{#time: [[Y]] m d }} → 2026 02 06

{{#time: [[Y (year)]] }} → 2026 (26UTCpmFri, 06 Feb 2026 13:35:33 +0000)

{{#time: [[Y "(year)"]] }} → 2026 (year)

{{#time: i's" }} → 35'33"

The date/time object can be in any format accepted by PHP's strtotime() function. Absolute (e.g. 20 December 2000), relative (e.g. +20 hours), and combined times (e.g. 30 July +1 year) are accepted.

{{#time: r|now}} → Fri, 06 Feb 2026 13:36:13 +0000

{{#time: r|+2 hours}} → Fri, 06 Feb 2026 15:36:13 +0000

{{#time: r|now + 2 hours}} → Fri, 06 Feb 2026 15:36:13 +0000

{{#time: r|20 December 2000}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|December 20, 2000}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|2000-12-20}} → Wed, 20 Dec 2000 00:00:00 +0000

{{#time: r|2000 December 20}} →

The language code in ISO 639-3 (?) allows the string to be displayed in the chosen language

{{#time:d F Y|1988-02-28|nl}} → 28 februari 1988

{{#time:l|now|uk}} → п'ятниця

{{#time:d xg Y|20 June 2010|pl}} → 20 czerwca 2010

The local parameter specifies if the date/time object refers to the local timezone or to UTC.

This is a boolean parameters: its value is determined by casting the value of the argument (see the official PHP documentation for details on how string are cast to boolean values).

See the following examples for details:

{{#time: Y F d H:i:s|now|it|0}} → 2026 febbraio 06 13:36:13

{{#time: Y F d H:i:s|now|it|1}} → 2026 febbraio 06 14:36:13

{{#time: Y F d H:i:s|+2 hours||0}} → 2026 February 06 15:36:13

{{#time: Y F d H:i:s|+2 hours||1}} → 2026 February 06 16:36:13

{{#time:c|2019-05-16T17:05:43+02:00|it}} → 2019-05-16T15:05:43+00:00

{{#time:c|2019-05-16T17:05:43+02:00|it|0}} → 2019-05-16T15:05:43+00:00

{{#time:c|2019-05-16T17:05:43+02:00|it|true}} → 2019-05-16T17:05:43+02:00

If you've calculated a Unix timestamp, you may use it in date calculations by pre-pending an @ symbol.

{{#time: U | now }} → 1770384973

{{#time: r | @1770384933 }} → Fri, 06 Feb 2026 13:35:33 +0000

Full or partial absolute dates can be specified; the function will "fill in" parts of the date that are not specified using the current values:

{{#time: Y | January 1 }} → 2026

A four-digit number is always interpreted as a year, never as hours and minutes:

{{#time: Y m d H:i:s | 1959 }} → 1959 02 06 00:00:00

A six-digit number is interpreted as hours, minutes and seconds if possible, but otherwise as an error (not, for instance, a year and month):

{{#time: Y m d H:i:s | 195909 }} → 2026 02 06 19:59:09 Input is treated as a time rather than a year+month code.

{{#time: Y m d H:i:s | 196009 }} → Although 19:60:09 is not a valid time, 196009 is not interpreted as September 1960.

The function performs a certain amount of date mathematics:

{{#time: d F Y | January 0 2008 }} → 31 December 2007

{{#time: d F | January 32 }} →

{{#time: d F | February 29 2008 }} → 29 February

{{#time: d F | February 29 2007 }} → 01 March

{{#time:Y-F|now -1 months}} → 2026-January

The total length of the format strings of the calls of #time is limited to 6000 characters.

Time Zone issue

There is a bug in this #time parser function (more specifically in PHP DateTime) that does not allow the passing-in of non-integers as relative time zone offsets. This issue does not apply when using an on-the-hour time zone, such as EDT. For example:

• {{#time:g:i A | -4 hours }} → 9:36 AM

However, India is on a +5.5 hours time offset from UTC, and thus using its time zone will not normally allow the correct calculation of a relative time zone offset. Here's what happens:

• {{#time:g:i A | +5.5 hours }} → 1:36 PM

To workaround this issue, simply convert the time into minutes or seconds, like this:

• {{#time:g:i A | +330 minutes }} → 7:06 PM
• {{#time:g:i A | +19800 seconds }} → 7:06 PM

(Tim Starling, the developer of this function, provided the exact syntax for this solution.)

#time format like in signatures

Sometimes it is useful to construct a timestamp, which looks like the automatic timestamp generated by signatures in discussions on talk pages. On an English-language wiki, it can be created with:

• {{#timel:H:i, j xg Y (e)|+330 minutes}} → 20:06, 6 February 2026 (Europe/Berlin)

#timel

This function is a syntactic shortcut that operates identically to {{#time: ... }} with the local parameter set to Template:Phpi, so it always uses the preferred time zone of the user or the configured time zone of the wiki (as set in Template:Ll)

Syntax of the function is:

{{#timel: format string }}

{{#timel: format string | date/time object }}

{{#timel: format string | date/time object | language code }}

Example of the use of #time and #timel parser functions from a server where the timezone is not UTC

For instance, see the following examples:

{{#time:c|now|it}} → 2026-02-06T13:36:13+00:00

{{#time:c|now|it|0}} → 2026-02-06T13:36:13+00:00

{{#time:c|now|it|1}} → 2026-02-06T14:36:13+01:00

{{#timel:c|now|it}} → 2026-02-06T14:36:13+01:00

Warning Example from https://no.wikipedia.org/wiki/Maldiskusjon:Sommertid

#timef

This function formats a date using a standard format for the selected language, as defined in $dateFormats (see Template:Phab).

{{#timef: date/time object }}

{{#timef: date/time object | format type }}

{{#timef: date/time object | format type | language code }}

The format of the date/time object is the same as for #time. If it is empty, the time when the page was rendered is used.

The format type may be one of:

time

Only the time is shown.

date

Only the date is shown.

both

Both the time and date are shown.

pretty

Only the date is shown, using an abbreviated format which does not include the year. Not all languages support this; if it is not supported, the "date" format is used.

If the format type is not specified, both the time and date will be show, as if both were specified.

If the language code is not specified, the page's content language is used.

Using #timef instead of #time allows templates to more easily support multiple languages, since different languages have different ways to format dates.

In English, the order of the day and month is controlled by Template:Wg.

Examples:

{{#timef:now|both|en}} → 13:36, 6 February 2026

{{#timef:now|both|ja}} → 2026年2月6日 (金) 13:36

{{#timef:now|pretty|en}} → 6 February

{{#timef:now|pretty|pl}} → 6 lutego

{{#timef:|time}} → 13:35

#timefl

This function is the same as #timef except that it uses the local timezone of the wiki as configured in Template:Ll.

{{#timefl: date/time object }}

{{#timefl: date/time object | format type }}

{{#timefl: date/time object | format type | language code }}

#titleparts

This function separates a page title into segments based on slashes, then returns some of those segments as output.

Template:Tmpl

If the number of segments to return parameter is not specified, it defaults to "0", which returns all the segments from the segment to start at to the end (included). If the segment to start at parameter is not specified or is "0", it defaults to "1":

{{#titleparts: Talk:Foo/bar/baz/quok }} → Talk:Foo/bar/baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | 1 }} → Talk:Foo See also {{Template:Ll}}.

{{#titleparts: Talk:Foo/bar/baz/quok | 2 }} → Talk:Foo/bar

{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 2 }} → bar/baz

{{#titleparts: Talk:Foo/bar/baz/quok | 2 | 3 }} → baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | 3 | 2 }} → bar/baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | | 2 }} → bar/baz/quok

{{#titleparts: Talk:Foo/bar/baz/quok | | 5 }} →

Negative values are accepted for both values. Negative values for the number of segments to return parameter effectively 'strips' segments from the end of the string. Negative values for the first segment to return translates to "start with this segment counting from the right":

{{#titleparts: Talk:Foo/bar/baz/quok | -1 }} → Talk:Foo/bar/baz Strips one segment from the end of the string. See also {{BASEPAGENAME}}.

{{#titleparts: Talk:Foo/bar/baz/quok | -4 }} → Strips all 4 segments from the end of the string

{{#titleparts: Talk:Foo/bar/baz/quok | -5 }} → Strips 5 segments from the end of the string (more than exist)

{{#titleparts: Talk:Foo/bar/baz/quok | | -1 }} → quok Returns last segment. See also {{SUBPAGENAME}}.

{{#titleparts: Talk:Foo/bar/baz/quok | -1 | 2 }} → bar/baz Strips one segment from the end of the string, then returns the second segment and beyond

{{#titleparts: Talk:Foo/bar/baz/quok | -1 | -2 }} → baz Start copying at the second last element; strip one segment from the end of the string

Before processing, the pagename parameter is HTML-decoded: if it contains some standard HTML character entities, they will be converted to plain characters (internally encoded with UTF-8, i.e. the same encoding as in the MediaWiki source page using this parser function).

For example, any occurrence of ", ", or " in pagename will be replaced by ".

No other conversion from HTML to plain text is performed, so HTML tags are left intact at this initial step even if they are invalid in page titles.

Then the decoded pagename is canonicalized into a standard page title supported by MediaWiki, as much as possible:

1. All underscores are automatically replaced with spaces:

   {{#titleparts: Talk:Foo/bah_boo|1|2}} → bah boo Not bah_boo, despite the underscore in the original.

2. The string is split a maximum of 25 times; further slashes are ignored and the 25th element will contain the rest of the string. The string is also limited to 255 characters, as it is treated as a page title:

   {{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee | 1 | 25 }} → y/z/aa/bb/cc/dd/ee

   If for whatever reason you needed to push this function to its limit, although very unlikely, it is possible to bypass the 25 split limit by nesting function calls:

   {{#titleparts: {{#titleparts: a/b/c/d/e/f/g/h/i/j/k/l/m/n/o/p/q/r/s/t/u/v/w/x/y/z/aa/bb/cc/dd/ee| 1 | 25 }} | 1 | 2}} → z

3. Finally the first substring is capitalized according to the capitalization settings of the local wiki (if that substring also starts by a local namespace name, that namespace name is also normalized).

   {{#titleparts: talk:a/b/c }} → Talk:A/b/c

String functions

The ParserFunctions extension optionally defines various string functions if $wgPFEnableStringFunctions is set to Template:Phpi:

• #len
• #pos
• #rpos
• #sub
• #count
• #replace
• #explode
• #urldecode

See the dedicated subpage for documentation, and Template:Ll for examples.

Here is a short overview of Module:String functions:

• #len (length of string): {{#invoke:String|len|target_string}}
• #sub (substring): {{#invoke:String|sub|target_string|start_index|end_index}}
• #match: {{#invoke:String|match|source_string|pattern_string|start_index|match_number|plain_flag|nomatch_output}}
• #pos (position of target): {{#invoke:String|pos|target_string|index_value}}
• #find: {{#invoke:String|find|source_string|target_string|start_index|plain_flag}}
• #replace: {{#invoke:String|replace|source_str|pattern_string|replace_string|replacement_count|plain_flag}}
• #rep (repeat): {{#invoke:String|rep|source|count}}
• #escapePattern: {{#invoke:String|escapePattern|pattern_string}}
• #count: {{#invoke:String|count|source_str|pattern_string|plain_flag}}
• #join: {{#invoke:String|join|separator|string1|string2|...}}

General points

Substitution

Parser functions can be substituted by prefixing the hash character with subst::

{{subst:#ifexist: Help:Extension:ParserFunctions/en | [[Help:Extension:ParserFunctions/en]] | Help:Extension:ParserFunctions/en }} → the code [[Help:Extension:ParserFunctions/en]] will be inserted in the wikitext since the page Help:Extension:ParserFunctions/en exists.

Substitution does not work within Template:Ll; you can use Template:((subst:#tag:ref|…Template:)) for this purpose.

Redirects

Especially {{#time:…|now-…}} could be handy in Template:Ll to pages including dates, but this does not work.

Escaping pipe characters

In tables

Parser functions will mangle Template:Ll syntax and pipe characters (|), treating all the raw pipe characters as parameter dividers. To avoid this, most wikis used a template Template:! with its contents only a raw pipe character (|), since MW 1.24 a Template:Ll replaced this kludge. This 'hides' the pipe from the MediaWiki parser, ensuring that it is not considered until after all the templates and variables on a page have been expanded. It will then be interpreted as a table row or column separator. Alternatively, raw HTML table syntax can be used, although this is less intuitive and more error-prone.

You can also escape the pipe character | for display as a plain, uninterpreted character using an HTML entity: | or | .

Description	You type	You get
Escaping pipe character as table row/column separator	{{!}}	|
Escaping pipe character as a plain character	|	|

In template calls

The same pipe protection applies as for the following example:

{{Documentation|content=
... text before ...<code>subpage-name=sandbox3</code> is equivalent to <code>sandbox link=../sandbox3 | sandbox name=sandbox3</code>.
... text after ...
}}

We observe that text after is not displayed when the pipe | just before sandbox name= is present since |sandbox name= is considered erroneously to be a parameter of template Documentation at the same level as |content= is.

Stripping whitespace

Whitespace, including newlines, tabs, and spaces, is stripped from the beginning and end of all the parameters of these parser functions. If this is not desirable, comparison of strings can be done after putting them in quotation marks.

{{#ifeq: foo           |           foo | equal | not equal }} → equal

{{#ifeq: "foo          " | "          foo" | equal | not equal }} → not equal

To prevent the trimming of then and else parts, see m:Template:If. Some people achieve this by using <Template:Ll> </nowiki> instead of spaces.

foo{{#if:|| bar }}foo → foobarfoo

foo{{#if:||<nowiki /> bar <nowiki />}}foo → foo bar foo

However, this method can be used to render a single whitespace character only, since the parser squeezes multiple whitespace characters in a row into one.

<span style="white-space: pre;">foo{{#if:||<nowiki/> bar <nowiki/>}}foo</span>

→

foo bar foo

In this example, the white-space: pre style is used to force the whitespace to be preserved by the browser, but even with it the spaces are not shown. This happens because the spaces are stripped by the software, before being sent to the browser.

It is possible to workaround this behavior replacing whitespaces with   (breakable space) or   (non-breakable space), since they are not modified by the software:

<span style="white-space: pre;">foo{{#if:||&#32;&#32;&#32;bar&#32;&#32;&#32;}}foo</span> → foo bar foo

foo{{#if:||   bar   }}foo → foo   bar   foo

Beware that not all parameters are created equal. In ParserFunctions, whitespace at the beginning and end is always stripped. In Template:Ll, whitespace at the beginning and end is stripped for named parameters and named unnamed parameters but not from unnamed parameters:

fooTemplate:Tlxfoo → fooTemplate:1xfoo

fooTemplate:Tlxfoo → fooTemplate:1xfoo

fooTemplate:Tlxfoo → fooTemplate:1xfoo

Other parser functions

Case conversion functions

• Lowercase: Template:Xpdoc
• Uppercase: Template:Xpdoc
• Lowercase first character: Template:Xpdoc
• Uppercase first character: Template:Xpdoc

Encoding functions

• URL encoding:

"{{urlencode: AbC
dEf ghi}}"

renders as

---

"AbC%0AdEf+ghi"

---

So inner new lines convert into %0A, and inner spaces convert into +.

Anchor encoding

{{anchorencode: AbC dEf ghi}}

renders as

---

AbC_dEf_ghi

---

Padding functions

• Template:Xpdop3p3c
• Template:Xpdop3p3c

Formatting functions

• Template:Xpd0

See also

• Template:Ll
• Template:Ll
• Template:Ll - an (incomplete) list of parser functions added by core and extensions.
• Template:Ll
• Template:Ll
• Template:Ll
• Module:String obsoleting Template:Ll
• Template:Ll
• Parser functions for Wikibase (the extensions that enables Wikidata): How to use data on Wikimedia projects#Parser function

References

[[Category:Extension help{{#translation:}}|ParserFunctions]] [[Category:Magic words{{#translation:}}]]